[Gta04-owner] Modem crashing?

[Dr. H. Nikolaus Schaller]

Am 15.02.2012 um 07:53 schrieb Dr. H. Nikolaus Schaller:

> 
> Am 15.02.2012 um 06:53 schrieb Radek Polak:
> 
>> On Tuesday 14 February 2012 13:52:54 Radek Polak wrote:
>> 
>>> Hi,
>>> i am now playing some more with GTA04's modem. Here is what i am doing:
>>> 
>>> 1/ Dial GTA04's number from other phone.
>>> 2/ Let it ring a few times.
>>> 3/ Hangup on the other phone.
>>> 
>>> After hangup i get this from kernel:
>>> 
>>> 	usb 1-2: USB disconnect, device number 2
>>> 	usb 1-2: new high-speed USB device number 3 using ehci-omap
>>> 	usb 1-2: New USB device found, idVendor=0af0, idProduct=8800
>>> 	usb 1-2: New USB device strings: Mfr=3, Product=2, SerialNumber=0
>>> 	usb 1-2: Product: Globetrotter HSUPA Modem
>>> 	usb 1-2: Manufacturer: Option N.V.
>>> 
>>> To me it looks like the modem crashed or was at least disconnected. Can
>>> someone try this too? It's under QtMoko v39. I will try from console now...
>> 
>> I found interesting thing - at home it never does it, but here at work it's 
>> 100% reproducible.
> 
> That is good to make it better tested.
> 
>> I think the problem can be that the modem switches to 3G at work.
>> 
>> Now can we somehow debug this more? Is there more documentation for the modem? 
>> E.g. can i switch from 3G to 2G only? Can i somehow use the Diagnostics ttyHS?
> 
> Unfortuantely I have no access to documentation of such details.
> 
>> 
>> Regards
>> 
>> Radek
>> 
>> 
>> 
>> Btw: it's also simply reproducible just using minicom:
>> 
>> AT+CFUN=1                                                                                                      
>> OK
>> AT+COPS
>> OK
>> 
>> RING		<-- here i started call
>> 
>> RING
>> 
>> RING
>> 
>>              <- here i hung up
>> 
>> ...and this was again in dmesg:
>> 
>> Feb 15 06:44:04 neo kernel: [  313.932647] 3G Wakeup
>> Feb 15 06:44:21 neo kernel: [  331.267059] usb 1-2: USB disconnect, device 
>> number 2
>> Feb 15 06:44:25 neo kernel: [  335.400665] usb 1-2: new high-speed USB device 
>> number 3 using ehci-omap
>> Feb 15 06:44:26 neo kernel: [  335.569732] usb 1-2: New USB device found, 
>> idVendor=0af0, idProduct=8800
>> Feb 15 06:44:26 neo kernel: [  335.578033] usb 1-2: New USB device strings: 
>> Mfr=3, Product=2, SerialNumber=0
>> Feb 15 06:44:26 neo kernel: [  335.586486] usb 1-2: Product: Globetrotter 
>> HSUPA Modem
>> Feb 15 06:44:26 neo kernel: [  335.592742] usb 1-2: Manufacturer: Option N.V.
> 
> Interesting issue. Do you run from battery in this mode? And do you have the device
> connected to the Freerunner antenna? Theoretically it could be some RF impulse
> or high current impulses that trigger a self-reset.
> 
> Another idea is to test if it really does reset the modem: please enable the SIM PIN.
> If you need to provide it again after this operation, the modem did reset. If not,
> it is just something in the USB interface. Maybe it is going to some suspend
> and waking up as a different device number?
> 
> One more idea: check with the 2.6.32 kernel and minicom. Then, we can
> exclude that it is a kernel/driver regression.
> 
> Finally it could be a defect on your board. I will try if I can reproduce it.

Yes, I have seen the same on a GTA04A3 board. Not every time but every
third or fourth.

at+clip=1
OK
at+cops
OK
at+crc=1
OK

+CLIP: "+498954290367",145,,,,0

+CRING: VOICE

+CLIP: "+498954290367",145,,,,0

+CRING: VOICE

+CLIP: "+498954290367",145,,,,0

+CRING: VOICE

+CLIP: "+498954290367",145,,,,0

+CRING: VOICE

+CLIP: "+498954290367",145,,,,0

dmesg:

[ 2523.762664] 3G Wakeup
[ 2587.356231] 3G Wakeup
[ 2619.660888] 3G Wakeup
[ 2795.349609] 3G Wakeup
[ 2819.874816] usb 2-2: USB disconnect, address 2
[ 2823.521575] usb 2-2: new high speed USB device using ehci-omap and address 3
[ 2823.695770] usb 2-2: New USB device found, idVendor=0af0, idProduct=8800
[ 2823.709136] usb 2-2: New USB device strings: Mfr=3, Product=2, SerialNumber=0
[ 2823.724700] usb 2-2: Product: Globetrotter HSUPA Modem
[ 2823.730041] usb 2-2: Manufacturer: Option N.V.

After that /dev/ttyHS3 has gone (minicom/femtocom is no longer responsive)
and the application port becomes /dev/ttyHS4.

Then, I did open a second ssh session and window and tracked dmesg.

The USB disconnect happens when the other side ends the ringing signal.
So it may indeed be a firmware bug. IMHO the GSM07.07 says that there
should be a NO CARRIER message.

What I also got when the first RING came:

[ 3197.607574] Spurious irq 95: 0xffffffdf, please flush posted write for irq 56

No idea what this means. Trouble at the 3G wakeup IRQ or some USB IRQ?

Nikolaus